UBC Theses and Dissertations
Bias and clinical judgment in counselling and psychotherapy: extending theory and research design Grigg, Donald G
In an investigation of how counsellors make decisions about how to intervene on behalf of their clients twelve trainees (six family counsellors and six person-centered counsellors) sorted a series of 32 possible counseling procedures from most to least desirable into a quasi-normal Q-distribution. Each subject completed the sort under four conditions, once for each hypothetical counselling client. The sex, age, gender, type of disorder, and severity of disorder were held constant across all conditions. Conditions varied as to socioeconomic status (poor versus middle class) and ethnicity (Caucasian versus Native) of the hypothetical client. The counselling alternatives ranked by the subjects had been pre-selected to represent two dimensions of cognitive attribution: internality versus externality and stability versus instability. The sorted cards yielded preference scores for each attributional category for each experimental condition. Analysis of variance was applied to the scores for each subject across all four conditions as a test for ethnic or socioeconomic status differences as a function of attributional pattern. The scores for all subjects were then correlated with each other, and this matrix of intercorrelations submitted to factor analysis yielding a principal axis solution. Varimax rotation was used to produce orthogonal factors. The factor loadings on each orthogonal factor were then used to calculate weighted standard scores indicative of the preference shown for each item within the pattern indicated by the factor. The items were than ranked by their standard scores, scored by attribution pattern, and analysis of variance applied. Analysis of variance for individual counsellors revealed only one subject for whom, across all conditions, ethnicity and socio-economic status were significant variables. Factor analysis revealed three distinct patterns of response, one pattern for family counsellors and two distinct patterns of response for the person-centered counsellors. Analysis of variance indicated that the family counsellors preferred items classified in the attributional schema as external or external-stable. The first group of person-centered therapists showed a preference for items classified as stable. The second group of person-centered therapists also preferred items classified as stable, but also systematically ranked unstable and external items lowest. Patterns were highly correlated across all experimental conditions except the "poor Caucasian" condition. In this condition a fourth pattern emerged representing only two person-centered therapists and showing no significant preferences on the attributional schema. The second person-centered pattern essentially disappeared, while the family therapists split into two less differentiated patterns. In the "poor Caucasian" condition there were four subjects whose responses correlated about equally with more than one pattern, something which did not occur in the other conditions. Data confirm that attribution theory and Q-methodology can be used to distinguish patterns of clinical decision-making as a function of type of therapist and client characteristics. Results also tend to confirm that clinical decision-making, as a pattern of attribution, is influenced by socio-economic status of the client.
Item Citations and Data